From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <micha-1@fantasymail.de>
X-Original-To: caml-list@yquem.inria.fr
Delivered-To: caml-list@yquem.inria.fr
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78])
	by yquem.inria.fr (Postfix) with ESMTP id AD142BB81
	for <caml-list@yquem.inria.fr>; Sun, 23 Oct 2005 23:37:56 +0200 (CEST)
Received: from mail.gmx.net (pop.gmx.de [213.165.64.20])
	by nez-perce.inria.fr (8.13.0/8.13.0) with SMTP id j9NLbuZD017232
	for <caml-list@yquem.inria.fr>; Sun, 23 Oct 2005 23:37:56 +0200
Received: (qmail invoked by alias); 23 Oct 2005 21:37:55 -0000
Received: from p54A30DFD.dip0.t-ipconnect.de (EHLO [192.168.2.136]) [84.163.13.253]
  by mail.gmx.net (mp019) with SMTP; 23 Oct 2005 23:37:55 +0200
X-Authenticated: #20477425
From: Michael Wohlwend <micha-1@fantasymail.de>
To: caml-list@yquem.inria.fr
Subject: Re: Another problem (was Re: [Caml-list] Ocamllex question)
Date: Sun, 23 Oct 2005 23:37:59 +0200
User-Agent: KMail/1.8
References: <435BD045.1000305@gushee.net> <435BFCD4.8040009@gushee.net>
In-Reply-To: <435BFCD4.8040009@gushee.net>
X-Face: S)[vu%Bha1d&ej9GfwAq~7C}A,y[B.uS}+D6'hb~xPwsxymw$fnCOaMe<=?utf-8?q?*bnUajSBR=5Fm=3FR=0A=09?=@V3;iX8[A}z`.%pEQ1r7iZhN8#ktTCBQ}&mkx>=RH&l|l6\]NZI@
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200510232337.59207.micha-1@fantasymail.de>
X-Y-GMX-Trusted: 0
X-Miltered: at nez-perce with ID 435C02B4.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)!
X-Spam: no; 0.00; caml-list:01 ocamllex:01 ocamllex:01 lexer:01 lazy:01 lexer:01 sequences:01 buffer:01 backslash:01 char:01 backslash:01 'n':01 escapes:01 buffer:01 char:01 
X-Spam-Checker-Version: SpamAssassin 3.0.3 (2005-04-27) on yquem.inria.fr
X-Spam-Level: 
X-Spam-Status: No, score=0.1 required=5.0 tests=FORGED_RCVD_HELO,SPF_FAIL 
	autolearn=disabled version=3.0.3

On Sunday 23 October 2005 23:12, Matt Gushee wrote:
> While we're on the subject of Ocamllex, there's another issue I'm
>
> wondering about. My lexer needs to handle quoted strings, something like:
>   | '"' [^ '"'] * as word '"'     { WORD word }
>
> Not essential, but it would be nice to allow escaped quotes within such
> strings:
>
>   "The quick brown fox jumped over the \"lazy\" dog."
>
> I've haven't actually tried to implement this yet, but thinking about it
> it seems like it would make the lexer hugely more complex. Can anyone
> suggest a reasonably simple way to deal with escape sequences?

you write a extra lexer-rule which gets called when a " is seen; in this rule 
you read single chars and append them to a string buffer. If you see a 
backslash you handle the next character special; example:

let char_for_backslash = function   
    | 'a' -> '\007'
    | 'v' -> '\011'
    | 'f' -> '\012'
    | 'n' -> '\n'
    | 't' -> '\t'
    | 'b' -> '\b'
    | 'r' -> '\r'
    | c   -> c

let bs_escapes = [ '\032' - '\255' ]

let string_buff = Buffer.create 256
let reset_string_buffer () = Buffer.clear string_buff  
let store_string_char c = Buffer.add_char string_buff c
let store_string s = Buffer.add_string string_buff s
let get_stored_string () = Buffer.contents string_buff

rule dict = parser
  | ...
   | [ '"'  ] as d 
        {
         reset_string_buffer(); 
         scan_str  lexbuf;
         let s = get_stored_string() in
         (* Printf.printf " String (%c) read:%s " d s; *)
         STRING(s) 
        }

and scan_str  = parse 
    | [ '"' ]  { () }
    | '\\' (bs_escapes as c)  
        { store_string_char (char_for_backslash c); scan_str lexbuf }
    | allowed_string_char as c {store_string_char c }
    | eof  { raise( Lexical_error("unterminated string") ) }


works well,

 Michael